Técnicas computacionais aplicadas para extração e centralização de informações dos protocolos clínicos e diretrizes terapêuticas

Abstract

Evidence-Based Medicine proposes a methodology for the medical decision-making process, minimizing the level of uncertainty and errors, as it does not only consider the professional's experience, but also the best available evidence about that disease. The Clinical Protocols and Therapeutic Guidelines (PCDT) are documents that contain evidence for an effective treatment and are developed by the National Commission for the Incorporation of Technologies in the Unified Health System (CONITEC). However, there is no centralization of information properly contained in a repository to facilitate consultation in a simplified and intuitive way. Thus, the proposal of the work was to implement, through computational text recognition techniques, a centralized repository with information present in PCDTs, seeking to promote a simplified, intuitive consultation and expanding access to information based on scientific evidence. The direct search technique is used to identify the position and extract the stretch of interest. The extracted snippet is stored in a database. The query for a search term query, entered by the user, will return which PCDTs have that term and their extracted content is returned to the user. The development of the work demonstrated that there is the possibility of centralizing the contents of interest present in the PCDT in order to facilitate the consultation in a simple and intuitive way. The information obtained can serve as a basis for application in improving the quality of care and the methods applied for extraction and consultation can also be applied to other PDF documents with the same purpose: centralization and facilitation of consultation